Documente Academic
Documente Profesional
Documente Cultură
NetApp Data ONTAP 8.1 operating in 7-Mode and Cluster-Mode, support in-place and nondisruptive expansion of existing 32-bit aggregates to the 64-bit format. 64-bit aggregates offer
a much larger size threshold while providing all the advantages and capabilities of aggregates,
including flexibility and storage efficiency. This technical report describes the 32-bit to 64-bit
expansion process and how to manage 64-bit aggregates in detail.
.
TABLE OF CONTENTS
1
BACKGROUND ..................................................................................................................................... 4
OVERVIEW ............................................................................................................................................ 4
2.1
2.2
BACKGROUND ......................................................................................................................................................... 6
3.2
3.3
3.4
REQUIREMENTS ...................................................................................................................................................... 8
3.5
CAVEATS .................................................................................................................................................................. 8
3.6
3.7
3.8
3.9
3.10
COMPRESSION ...................................................................................................................................................... 14
4.2
FLEXCLONE ........................................................................................................................................................... 14
4.3
4.4
PERFORMANCE.................................................................................................................................. 21
5.1
5.2
5.3
5.4
CONCLUSION ...................................................................................................................................... 24
REFERENCES ..................................................................................................................................... 25
8.1
8.2
LIST OF TABLES
Table 1) Maximum aggregate and FlexVol volume sizes. ............................................................................ 6
Table 2) Block pointer size in 32-bit and 64-bit aggregates. ........................................................................ 6
Table 3) Command-set summary ............................................................................................................... 11
Table 4) Overview of various data replication features in Data ONTAP 8.1 7-Mode and Cluster-Mode. .. 14
Table 5) SnapMirror supportability matrix for 7-Mode. ............................................................................... 15
Table 6) SnapMirror supportability matrix for Cluster-Mode. ...................................................................... 16
Table 7) Interoperability with Data ONTAP features .................................................................................. 19
Table 8) Sample completion times for certain configurations. .................................................................... 24
LIST OF FIGURES
Figure 1) Existing aggregates on the storage system show up as 32-bit aggregates after upgrading to
Data ONTAP 8.1. .......................................................................................................................................... 5
Figure 2) Default format for new aggregates will be 64-bit. .......................................................................... 5
Figure 3) 32-bit to 64-bit expansion .............................................................................................................. 7
Figure 4) 64-bit to 32-bit SnapMirror ........................................................................................................... 17
Figure 5) 32-bit to 64-bit SnapMirror ........................................................................................................... 17
Figure 6) Data replication utilities and their interaction with 32-bit and 64-bit aggregates. ........................ 20
Figure 7) Throughput sustained during 64-bit expansion. .......................................................................... 22
Figure 8) Latency measurement during 64-bit expansion. ......................................................................... 22
Figure 9) Throughput sustained during 64-bit expansion. .......................................................................... 23
Figure 10) Latency measurement during 64-bit expansion. ....................................................................... 23
1 BACKGROUND
Aggregates provide storage to the volumes they contain. Each aggregate has a set of assigned disks or
array LUNs. An aggregate can hold one or more FlexVol volumes. These FlexVol volumes share the
physical resources of the underlying container.
Aggregates can be in 32-bit or 64-bit block format. A FlexVol volume inherits the parent aggregates
format. 32-bit aggregates are limited in size to 16TB. The 16TB limit is an artifact of the maximum
addressable space with a 32-bit block pointer in a 32-bit aggregate.
64-bit aggregates contain larger block pointers (64-bit) and therefore can grow beyond 16TB. Hence, they
have larger size limits than 32-bit aggregates.
In Data ONTAP 8.0, users could create 32-bit or 64-bit aggregates. Both aggregate types can coexist in a
system. However, Data ONTAP 8.0 was still subject to the following limitations:
An administrator wishing to expand any of the existing 32-bit aggregates to 64-bit is required to copy
out the data.
Existing 32-bit aggregates are unable to grow beyond 16TB seamlessly.
Physical SnapMirror replication engines are restricted to volumes residing in the same aggregate
type.
All of the above limitations have been fixed in Data ONTAP 8.1 which makes the aggregate format and
the expansion process seamless to the system administrator. The administrator can now expand existing
32-bit aggregates in place without requiring any data migration.
2 OVERVIEW
Data ONTAP 8.1 provides the following new functionality:
By default, all newly created aggregates (including the root aggregate) in Data ONTAP 8.1 will now
be in 64-bit format.
A 32-bit aggregate and its FlexVol volumes are automatically expanded to 64-bit without requiring a
data copy when new disks are added to push the total aggregate size beyond 16TB. There is no
disruption to client data access during this process.
Physical SnapMirror relationships work across volumes residing in any aggregate type.
Maximum size limits for aggregates have been increased.
One of the key new features is the functionality to expand existing 32-bit aggregates in place, that is,
without requiring any data migration or data copy. The expansion process is non-disruptive and occurs
while the volumes are online and serving data to the clients. The expansion process is independent of the
Data ONTAP upgrade procedure and it works seamlessly with all existing Data ONTAP features. This
feature is available in Data ONTAP operating in both 7-Mode and Cluster-Mode.
Note:
2.1
Unless explicitly stated all content is applicable to both 7-Mode and Cluster-Mode.
All new aggregates are by default 64-bit aggregates irrespective of the size of the aggregate. This also
applies to the root aggregate. An upgrade to Data ONTAP 8.1 does not change the format of any existing
aggregates. The administrator will have to add disks to trigger the 64-bit expansion process.
SCENARIO 1: UPGRADE TO DATA ONTAP 8.1
On an upgrade to Data ONTAP 8.1, there are no changes. All aggregates will continue to retain the same
format. All existing 32-bit aggregates will continue to be 32-bit.
Figure 1) Existing aggregates on the storage system show up as 32-bit aggregates after upgrading to Data
ONTAP 8.1.
Aggr
[32-bit]
Aggr
Aggr
[32-bit]
[32-bit]
(Root)
Aggr
[32-bit]
(Root)
Aggr
[64-bit]
Aggr
[64-bit]
NetApp storage
system running Data
ONTAP 8.0.x
NetApp storage
system running Data
ONTAP 8.1
Aggr
[64-bit]
Aggr
[64-bit]
(Root)
Aggr
[64-bit]
NetApp storage
system running Data
ONTAP 8.1
2.2
Maximum size limits for aggregates have been increased in Data ONTAP 8.1 and vary based on the
platform. Volume sizes stay the same as in Data ONTAP 8.0.
Table 1) Maximum aggregate and FlexVol volume sizes.
FAS/V-Series
Maximum Aggregate
Size
Maximum Aggregate
Size
Platform
8.1
8.0
(TB)
(TB)
162
100
100
6210
162
70
70
105
70
70
3240, 3160
90
50
50
3210, 3140
75
50
50
3070, 3040
50
50
50
2040
50
30
30
Note:
BACKGROUND
A file is made up of individual data blocks and large files have additional layers of indirection between the
inode and the actual data blocks. These additional layers are referred to as indirect blocks.
The block addresses for 64-bit aggregates are stored in 64 bits. The block addresses for 32-bit
aggregates are stored in 32-bit. With the increased block pointer size in 64-bit aggregates, the number of
addresses or block pointers that an indirect block can hold has decreased. Consequently, this results in
an inode requiring more indirect blocks in a 64-bit volume or aggregate.
Table 2) Block pointer size in 32-bit and 64-bit aggregates.
32-bit aggregates
64-bit aggregates
32-bit
64-bit
510
255
3.2
The expansion process expands all existing 32-bit indirect blocks within the aggregate and its containing
FlexVol volumes to the 64-bit format. It walks through all the indirect blocks in every inode tree, allocating
additional blocks when needed to hold the larger 64-bit pointers. The expansion process rewrites only
indirect blocks and does not touch the data blocks.
Once the expansion process is initiated all new writes will always write out blocks in the 64-bit format.
Figure 3) 32-bit to 64-bit expansion
Online
Non-disruptive
INSTALLATION/LICENSE
No special license is required to enable this feature.
3.3
The expansion process is triggered only when you add disks to a 32-bit aggregate such that the resulting
aggregate capacity is greater than 16TB. When disks are added, if the resulting capacity is less than
16TB, the aggregate will continue to be 32-bit.
How to trigger the expansion process
Add enough storage to a 32-bit aggregate to increase the size beyond 16TB.
For best performance, add a complete RAID group to prevent the new disks from becoming a
performance bottleneck as all new client writes are directed to the newly added disks.
When disks are added, Data ONTAP checks that there is enough free space for the extra metadata
blocks required for the 64-bit expansion. If all the volumes have sufficient free space, the disks are added
and the aggregate is expanded to 64-bit. If any of the volumes are too full to accommodate the 64-bit
expansion, the command will fail.
64-BIT EXPANSION WORKFLOW
The administrator would execute the following sequence of steps to expand 32-bit aggregates.
1. Establish requirement to grow an aggregate beyond 16TB.
2. Add enough disks to grow the aggregate beyond 16TB.
3. Background expansion starts on all volumes (excluding read-only SnapMirror destinations) within the
aggregate. All new writes will be in 64-bit format.
4. Expansion process finishes on all volumes within the aggregate.
Clients can continue to access their data during the entire process.
3.4
REQUIREMENTS
Disks must be added to an existing 32-bit aggregate such that the aggregates storage capacity
exceeds 16TB. This is the only action that is available to customers to trigger the expansion process.
Enough free space should be available to enable completion of the expansion since 64-bit
aggregates consume more metadata. See section 3.6 for more details.
Expansion on any 32-bit aggregate created in prior releases can be triggered only after the
completion of the prequalification scanner. The prequalification scanner is triggered on an upgrade to
Data ONTAP 8.1. See section 3.6 for more details.
3.5
CAVEATS
If the 32-bit aggregate you wish to expand contains any FlexCache volumes, the cache volumes will
have to be destroyed. The cache volumes can be recreated once the expansion process has started.
This causes FlexCache clients to lose their existing connections. These clients must reconnect to the
newly created FlexCache volumes and may experience an initial slowdown in data access due to
cache repopulation.
DataMotion for Volumes for 7-Mode volumes will not work across 32-bit and 64-bit aggregates.
DataMotion for Vfilers (7-Mode only) will not work across 32-bit and 64-bit aggregates.
Note:
DataMotion for Volumes in Cluster-Mode does not have the same restriction. Cluster-Mode
volumes can be moved between 32-bit and 64-bit aggregates.
Limitations
FlexCache volumes need to be destroyed before the expansion process can be triggered.
DataMotion for Volumes in 7-Mode will not work between 32-bit and 64-bit aggregates.
DataMotion for Vfilers (7-Mode only) will not work across 32-bit and 64-bit aggregates.
3.6
SPACE CONSUMPTION
64-bit aggregates consume additional metadata with the increase in block pointer size. Hence, the
expansion process will need to allocate extra blocks. If the volume is approaching its full capacity, it is
possible to run out of space during the expansion process.
If your volume(s) are close to full, you can increase the size of the volumes beforehand. When triggering
the expansion process, the administrator will be notified about the list of volumes that will require
additional space for the expansion process.
PREQUALIFICATION SCANNER
Because the expansion process requires additional space, a prequalification process is triggered on an
upgrade to Data ONTAP 8.1 to gather information on all the space-reserved files in existing volumes so
that we can continue to enable space reservations. This scanner provides a quick estimate of additional
blocks needed to expand a volume. The expansion process uses this estimate to determine if there is
sufficient space in the volumes to successfully trigger the expansion and maintain space guarantees. As
such the expansion process can only be triggered after the completion of the prequalification phase. If the
expansion process is triggered in the midst of the pre-qualification phase, it will error out with a more
detailed message.
The scanner is not triggered for any new volumes created in Data ONTAP 8.1 as the required space
accounting is maintained from the beginning.
I INCREASED THE VOLUME SIZE BUT THE EXPANSION PROCESS IS STILL OUT OF SPACE.
Even if space was added to a volume before the expansion process, it is still possible to run out of space
since client traffic may consume additional space. This scenario is less likely since the expansion is
triggered by the addition of disks, which results in more usable capacity.
WHAT HAPPENS IF I RUN OUT OF SPACE DURING THE EXPANSION?
If the volume is out of space during the expansion process, the process aborts. Clients may start seeing
ENOSPC errors. The appropriate volume autogrow/autodelete configurations will be triggered.
Meanwhile, the administrator can create more space by either:
1. Growing the volume or
2. Adding more disks
The expansion process is restarted automatically when space is available.
WHAT HAPPENS TO MY QUOTAS?
If the quotas are close to the limit, it is possible to exceed quotas during the expansion process due to
additional blocks being required.
3.7
Status
raid_dp, aggr
32-bit
Options
Status
raid_dp, aggr
32-bit
Options
elis22b> df -A aggr1
Aggregate
kbytes
used
avail capacity
aggr1
14037131520 2159414748 11877716772
15%
STEP 4: ADD DISKS WITH THE 64-BIT OPTION TO TRIGGER THE EXPANSION
The administrator will now be required to specify the -64bit-upgrade option when adding disks to trigger
the 64-bit expansion.
elis22b> aggr add aggr1 -64bit-upgrade normal 12
Note: preparing to add 10 data disks and 2 parity disks.
Continue? ([y]es, [n]o, or [p]review RAID layout) y
File system size 18.56 TB exceeds maximum 15.99 TB
Checking for additional space required to upgrade all writable 32-bit
volumes in aggregate aggr1 (Ctrl-C to interrupt).....
Addition of 12 disks to the aggregate has completed.
[elis22b:wafl.scan.64bit.upgrade.start:notice]: The 64-bit upgrade scanner has
running on volume vol1.
[elis22b:wafl.scan.64bit.upgrade.start:notice]: The 64-bit upgrade scanner has
running on aggregate aggr1.
[elis22b:wafl.scan.64bit.upgrade.completed:notice]: The 64-bit upgrade scanner
completed running on volume vol1.
[elis22b:wafl.scan.64bit.upgrade.completed:notice]: The 64-bit upgrade scanner
completed running on aggregate aggr1..
10
started
started
has
has
If any of the volumes are too full, the command will error out. The administrator can run the space check
command aggr add aggr1 -64bit-upgrade check to get more information on the list of volumes that
need to be grown and the amount of space that needs to be added to each of those volumes.
STEP 5: AGGREGATE IS NOW IN 64-BIT FORMAT
As soon as the expansion process is triggered, the aggregate will be considered to be a 64-bit aggregate.
Even if the process is not finished yet, the aggregate is a 64-bit aggregate with a mix of 32-bit and 64-bit
indirect blocks.
elis22b*> aggr status aggr1
Aggr State
aggr1 online
Status
raid_dp, aggr
parity uninit'd!
64-bit
Options
3.8
-all
% Completed
6
Time to Completion
2976
Progress
fbn 0,
10
1719
fbn 0,
10
1736
fbn 0,
COMMAND-SET SUMMARY
64-bit expansion is supported in Cluster-Mode too and the behavior is the same as in 7-Mode. The
following table lists the entire command set required for 64-bit expansion.
Table 3) Command-set summary
11
Feature
7-Mode
Cluster-Mode
3.9
When total space after disk selection crosses the 16TB limit for a 32-bit aggregate, System Manager will
warn the administrator to expand to the 64-bit format. On confirmation, System Manager will launch a
wizard to help the administrator expand the aggregate.
STEP 1: ADD DISKS TO GROW AN AGGREGATE (SIZE >16TB)
12
13
COMPRESSION
Compression is not supported on 32-bit volumes. Compression can be enabled as soon as you trigger
the 32-bit to 64-bit expansion process on the aggregate.
Best Practice
Enable compression after the completion of the expansion process.
4.2
FLEXCLONE
If the aggregate that is in the midst of being expanded contains any FlexClone volumes, the 64-bit
expansion process will expand the indirect blocks in the FlexClone volume too. This will cause parts of
the FlexClone volume to be split from its parent because it will overwrite all indirect blocks shared with the
parent volume.
4.3
DATA REPLICATION
Table 4 gives an overview of the various replication engines in 7-mode and Cluster-mode.
Table 4) Overview of various data replication features in Data ONTAP 8.1 7-Mode and Cluster-Mode.
Feature
7-Mode
Cluster-Mode
Works across
32-bit and 64bit
aggregates?
replication
Asynchronous Volume
SnapMirror
Yes
Physical
Yes
Physical
Synchronous Volume
SnapMirror
Yes
Physical
N/A
Not supported
Aggr Copy
Yes
Physical
N/A
Not supported
Vol Copy
Yes
Physical
Yes
Physical
DataMotion for
Volumes
No
Physical
Yes
Physical
Load-Sharing Mirror
N/A
Not supported
Yes
Physical
14
Nature of
Nature of
replication
7-MODE
In Data ONTAP 8.0 operating in 7-Mode, all physical-based replication engines like volume SnapMirror,
Aggr Copy, and Vol Copy are restricted to aggregates and volumes of the same type. Starting in Data
ONTAP 8.1, you can create physical-based replication relationships between volumes of different
formats.
Please refer to the following supportability matrix for all supported SnapMirror relationships across
different formats and Data ONTAP kernels for 7-Mode.
Table 5) SnapMirror supportability matrix for 7-Mode.
Destination Aggregate
Source
Volume
7.3.x
(32-bit)
8.0.x
(32-bit)
8.0.x
(64-bit)
8.1.x
(32-bit)
8.1.x
(64-bit)
7.3.x
(32-bit)
Yes
Yes
No
Yes
Yes
8.0.x
(32-bit)
No
Yes
No
Yes
Yes
8.0.x
(64-bit)
No
No
Yes
Yes
Yes
8.1
(32-bit)
No
No
No
Yes
Yes
8.1
(64-bit)
No
No
No
Yes
Yes
CLUSTER-MODE
Existing 10.0.x and 8.0.x mirror relationships must be deleted and new baseline transfers must occur as
part of the upgrade process to Data ONTAP 8.1 operating in Cluster-Mode.
Unlike, Data ONTAP 8.0 operating in 7-Mode, the async replication engine in Data ONTAP 8.0 operating
in Cluster-Mode can replicate between a 32-bit and a 64-bit volume.
Starting in Data ONTAP 8.1, you can create physical-based replication relationships between volumes of
different formats irrespective of the mode.
Please refer to the following supportability matrix for all the supported SnapMirror relationships across
different formats and Data ONTAP kernels for Cluster-Mode.
15
Destination Aggregate
Source
Volume
10.0.x
(32-bit)
8.0.x
(32-bit)
8.0.x
(64-bit)
8.1.x
(32-bit)
8.1.x
(64-bit)
10.0.x
(32-bit)
Yes
No
No
No
No
8.0.x
(32-bit)
No
Yes
Yes
No
No
8.0.x
(64-bit)
No
Yes
Yes
No
No
8.1
(32-bit)
No
No
No
Yes
Yes
8.1
(64-bit)
No
No
No
Yes
Yes
16
64-bit Aggr
32-bit Aggr
64-bit source
64-bit
destination
Flexvol
Volume SnapMirror
Flexvol
2. It is also possible to create a SnapMirror relationship from a 32-bit volume to a volume in a 64-bit
aggregate. This will result in having a 32-bit destination volume residing in a 64-bit aggregate.
32-bit Aggr
64-bit Aggr
32-bit source
32-bit
destination
Flexvol
17
Volume SnapMirror
Flexvol
Note:
A SnapMirror destination is skipped when the aggregate is expanded to 64-bit. The 32-bit
SnapMirror destination will be expanded to 64-bit only on a SnapMirror break or if the source gets
expanded to 64-bit.
SnapMirror flip resync
Consider the above scenario in which a 32-bit source volume is mirrored with SnapMirror to a
volume in a 64-bit aggregate. 64-bit expansion is triggered on the 32-bit destination volume when
the SnapMirror relationship is broken. If the administrator attempts to do a flip resync from the
destination volume back to the original 32-bit source volume, the original 32-bit source volume
will be converted to a 64-bit volume in a 32-bit aggregate.
18
4.4
Feature
Compatibility
Qtree
SnapMirror/SnapVault
(Supported in 7-Mode only)
NDMP
No restriction. It can be triggered during the expansion process. The data can
be restored to any format volume.
Synchronous SnapMirror
(Supported in 7-Mode only)
Asynchronous SnapMirror
SnapLock
(Supported in 7-Mode only)
RAID
SyncMirror/MetroCluster
Vol Copy
Aggr Copy
(Supported in 7-Mode only)
Compression
Deduplication
It can be triggered during the expansion process. It does not work across
aggregate types.
It can be triggered during the expansion process and also works between 32bit and 64-bit volumes.
FlexCache
Reallocate
No restrictions.
FlexClone
SnapRestore
Volume transition
Autosize
Any autosize changes are not honored until the completion of the expansion
process.
19
Feature
Compatibility
Failover/Failback
Figure 6) Data replication utilities and their interaction with 32-bit and 64-bit aggregates.
32-bit
32-bit
Aggr
Aggr
FlexVol
All
FlexVol
All except
DataMotion for
Volumes in 7-Mode
64-bit
64-bit
Aggr
Aggr
FlexVol
All
FlexVol
All
NetApp
Storage
System
20
NetApp
Storage
System
5 PERFORMANCE
5.1
For client operations on inodes that are not yet fully expanded, additional block lookups are sometimes
required to perform the 64-bit expansions inline. Since the client operations may also trigger the 64-bit
expansion, this may increase the latency and reduce throughput. Once an inode is fully expanded to 64bit, the extra block lookups are not required. The performance has the highest hit when the expansion is
first triggered, then improves gradually as the expansion process progresses.
PERFORMANCE IMPACT ON SNAPMIRROR DESTINATIONS
If an aggregate containing a 32-bit SnapMirror destination volume is expanded to 64-bit before the 32-bit
source aggregate, read throughput may be impacted on the destination.
This problem only exists until the source volumes are expanded to 64-bit format and the new blocks are
transferred to the destination. Until then, there can be a performance hit for read workloads due to the
extra lookups required to resolve physical addresses. The performance degradation for sequential read
workloads on such destinations is no more than 15% for various datasets in the worst case.
DOES THE TIME TAKEN FOR THE EXPANSION PROCESS DEPEND ON THE SIZE OF THE
AGGREGATE/VOLUMES?
No. The expansion process updates the format of the indirect metadata in the aggregate and the volume.
So, it depends on the amount of indirect metadata.
WHAT PERCENTAGE OF THE DATASET COMPRISES INDIRECT BLOCKS?
The amount of indirect metadata can vary based on the number of files and the amount of used data.
Typically, it accounts for 2% or less of the dataset.
5.2
OLTP BENCHMARK
SPC-1 is a standard benchmark that generates a workload with characteristics of typical business
applications such as database and e-mail with random I/O, queries, and updates.
The baseline test was performed with 32-bit aggregates while simulating the SPC-1 benchmark. To
measure the impact of the 64-bit expansion process on system latency, the same test was repeated while
sustaining the same throughput as the baseline and triggering the expansion process simultaneously.
21
25000
20000
15000
10000
Baseline:total_ops
64-bit expansion:total_ops
5000
0
1
Latency (ms)
Baseline:sys_avg_latency
64-bit
expansion:sys_avg_latency
1
5.3
FILE-SHARING WORKLOADS
The SFS2008 NFS benchmark allows customers to gauge home directory performance with an industrystandard tool. The following section provides performance data while simulating the SFS2008 NFS
benchmark with the 64-bit expansion process. The test measures the latency impact of the 64-bit
22
expansion process on client traffic while trying to maintain the same throughput as the baseline run with
32-bit aggregates.
The results contain 9 load points and the results of the server at each of the requested load points. Each
load point presents a requested number of operations per second.
Figure 9) Throughput sustained during 64-bit expansion.
50000
40000
30000
20000
Baseline:total_ops
10000
64-bit expansion:total_ops
0
1
Latency (ms)
5
4
3
Baseline:sys_avg_latency
2
64-bit
expansion:sys_avg_latency
1
0
1
23
5.4
COMPLETION TIME
Table 8 can be used as a guideline for how long the expansion process might take to complete on an idle
system. Several factors determine the timing, such as system configuration, I/O workload, size of dataset
in terms of inodes and blocks, nature of dataset, and so on.
Table 8) Sample completion times for certain configurations.
Aggregate Size
Volume Size
Number of
Volumes
Usage in Each
Volume
Completion
Time
Config 1
16TB
10G
499
50
~30 hrs
Config 2
16TB
13TB
100
~28 hrs
Config 3
13TB
12TB
30
~10 hrs
Config 4
13TB
1TB
10
90
~1 hr
Config 5
13TB
100G
100
70
~1 hr
Config 6
13TB
25G
499
66
~1 hr
Config 7
13TB
400G
40
30
~1 hr
7 CONCLUSION
Data ONTAP 8.1 operating in 7-Mode and Cluster-Mode gives the administrator the flexibility to
seamlessly expand existing 32-bit aggregates to 64-bit aggregates without any disruption to client access.
These enhancements make the format of the aggregate (32bit or 64bit) largely invisible to the customer.
24
8 REFERENCES
8.1
8.2
http://now.netapp.com/NOW/knowledge/docs/ontap/ontap_index.shtml
NetApp provides no representations or warranties regarding the accuracy, reliability or serviceability of any
information or recommendations provided in this publication, or with respect to any results that may be
obtained by the use of the information or observance of any recommendations provided herein. The
information in this document is distributed AS IS, and the use of this information or the implementation of
any recommendations or techniques herein is a customers responsibility and depends on the customers
ability to evaluate and integrate them into the customers operational environment. This document and
the information contained herein may be used solely in connection with the NetApp products discussed
in this document.
Go further, faster
Copyright 2012 NetApp, Inc. All rights reserved. No portions of this document may be reproduced without prior written consent of
NetApp, Inc. Specifications are subject to change without notice. NetApp, the NetApp logo, Go further, faster, xxx, and xxx are
trademarks or registered trademarks of NetApp, Inc. in the United States and/or other countries. All other brands or products are
trademarks or registered trademarks of their respective holders and should be treated as such. TR-3978-0911
25